"The desktop" and singularity (also scanning and ferris)

While I normally don't waffle on about higher level touchy feely sorts of things, a recent acquisition of a new scanner prompted my mind to wander through those neurological paths. I think it's fair to say that many folks don't run just one, but many desktops these days. And no, this is not about KDE vs GNOME vs emacs as your desktop, but rather that a whole KDE4 session is happening on another (virtual) machine which is brought across to the normal "main" desktop.

This gives me many apps running on many machines and also proxy X sessions and X-VNC sessions in windows (which normally have an embedded panel et al). A take away idea I had here was why when I have Abiword in the menus is there not an offer to run it on server-Y which I have an ssh key for instead of the local host? Or to bring xpra into the mix there as well. Sure, this is likely not an idea that folks who only own a single machine or laptop will like, but that demographic is surely shrinking fast. Add in nepomuk for tracking this and if I run amarok on furryshark7 most of the time then surely that can become the default host for it's icon rather than localhost.

The scanner idea was simple, a USB scanner, a USB hub to attach it to the host through and a handful of USB sticks. Each stick has a profile (username, colour, dpi etc) and to scan just put the document into the scanner and drop in the usb stick for a moment. Then the host knows the parameters and can put the file into a user specific directory on an intranet share so only the document holder can actually see the scanned version. Privacy and convenience, a rare thing!

Of course now that I have a scanner which doesn't suck, libferris will be getting sane support so I can "cp" documents right from the scanner to the filesystem or web...

cp sane://my-scanner-model/600dpi/colour/adf1.jpg flickr://me/upload

And the reason files are numbered is so a single copy can grab 30 documents from the ADF with a normal copy syntax. In practice, adf2 and adf1 both just scan the next piece of physical paper and give you it encoded in an appropriate manner (jpg for example). The fun part will be smoothing over a paper jam at document 7 of 16.

sync && umount blog://